GaRAGe: A Benchmark with Grounding Annotations for RAG Evaluation
This provides a benchmark for fine-grained evaluation of RAG systems, addressing a specific need in the AI/ML community for assessing grounding accuracy in real-world scenarios.
The authors tackled the problem of evaluating LLMs' ability to identify relevant grounding passages in RAG systems, finding that models often over-summarize rather than strictly ground answers, with scores like at most 60% for relevance-aware factuality and 58.9% F1 in attribution.
We present GaRAGe, a large RAG benchmark with human-curated long-form answers and annotations of each grounding passage, allowing a fine-grained evaluation of whether LLMs can identify relevant grounding when generating RAG answers. Our benchmark contains 2366 questions of diverse complexity, dynamism, and topics, and includes over 35K annotated passages retrieved from both private document sets and the Web, to reflect real-world RAG use cases. This makes it an ideal test bed to evaluate an LLM's ability to identify only the relevant information necessary to compose a response, or provide a deflective response when there is insufficient information. Evaluations of multiple state-of-the-art LLMs on GaRAGe show that the models tend to over-summarise rather than (a) ground their answers strictly on the annotated relevant passages (reaching at most a Relevance-Aware Factuality Score of 60%), or (b) deflect when no relevant grounding is available (reaching at most 31% true positive rate in deflections). The F1 in attribution to relevant sources is at most 58.9%, and we show that performance is particularly reduced when answering time-sensitive questions and when having to draw knowledge from sparser private grounding sources.